← Back to the exerciseThe full worked example: Who should own renewals? Compare three proposals on the same measure
campo.
Field guide / Sales / UC-003 · SC-05
The result

Who owns renewals? None of the three proposals is proven. By dollars, the pilot kept 85% against sales' 88%.

A CRO at a 58,000-person industrial automation company compares three proposals for who owns renewals before she meets the CEO and CFO, with any AI chat her company allows.

The resultHelena's page for the CEO and CFO. Fictional company.
TALVORA GROUP · REVENUE
Who owns renewals: recommendation
For: Andrew Hale, CEO, and Sofia Brandt, CFO
From: Helena Vasquez · Tuesday, March 2

MY RECOMMENDATION

Don't approve a change on the current evidence. None of the three proposals is shown to beat today's 88% of renewal dollars, and B and C each cost far more than they state while account executives are still paid on renewals. Renewals stay with sales at today's pay until revenue operations compares like with like. · Inputs 2, 3, 4

THE THREE OPTIONS (EXTRA RENEWAL PAY, FULL-YEAR ESTIMATE, AGAINST TODAY)

A: sales keeps+$1.3M a year
Full year; the 92% has no basis
B: customer success+$3.8M a year
While AEs keep 2% (2027 plan); then $1.5M less; pilot kept 85% of dollars
C: renewals team+$5.5M a year
While AEs keep 2% (2027 plan); then +$0.2M; 94% is a vendor benchmark
CustomersSecond-hand only
No direct input; the 60 largest are $132M

WHAT THE COMPARISON SHOWS, AND WHAT'S NEXT

SAME MEASURE
"95% against 88%" compares contracts with dollars
By contracts, the pilot kept 95% and sales 91%. By dollars, 85% and 88%. The pilot was six months of contracts under $250K, with none of the 60 largest. · Input 3
COST BASIS
B and C pay the same renewal twice while account executives keep their 2%
Costs are full-year estimates using 2026 renewed revenue ($263.6M). They include stated salaries, commissions and bonuses, but exclude benefits, hiring and other costs. Actual 2027 spending depends on when changes take effect; HR and the CFO confirm. · Input 4
BY FRI MAR 5
Revenue operations: sales' results for the pilot's kind of contracts, in contracts and dollars
Plus what our customers want, starting with the 60 largest. Today we only have it second-hand. · Inputs 3, 5
Caught on review
1
"95% against 88%" compares contracts with dollars: by dollars the pilot kept 85%
2
Options B and C leave out pay: in 2027 the same renewal is paid twice
3
The 92% and 94% have no support in our own data
UC-003 · SC-05 · Rev B
FOR
Chief Revenue Officer
SECTOR
Industrial automation (software and service contracts)
COMPANY
58,000 people · 40 sites
TASK
Recommend who should own renewals, from three competing proposals
WRITTEN FOR
ChatGPT · Claude · Gemini · Copilot
AI TEST
Claude, Nov 15, 2026 · 7 of 8 found, 1 partly
Step 0 · The mess

What landed on Helena's desk

A short email from the CEO, three one-line proposals, last year's renewal data, HR's pay notes and a note from the VP Customer Success. Every proposal has a strong number. None says what it counts.

EMAIL · CEO[1]
Who owns renewals
“Renewals slipped to 88% last year … one recommendation on March 11, with what it costs.”
1
The decision.
THREE PROPOSALS[2]
A
“Our pilot retained 95% against sales' 88%.” Cost stated: $1.12M a year.
2
95% of what?
REVENUE OPS[3]
Renewal data 2026
Pilot: 100 contracts, each under $250K, all due July to December.
3
Same accounts?
HR COMPENSATION[4]
Pay notes
“Account executives keep their 2% for 2027.”
4
Paid twice.
NOTE · VP CUSTOMER SUCCESS[5]
Customers
“The pilot proves it.”
5
Who asked customers?
You're not alone
3%
of B2B software leaders surveyed by McKinsey said they have a best-in-class coverage model for customer success and renewals.
McKinsey, The net revenue retention advantage, November 2025: more than 100 leaders at 98 B2B software companies. Source
Steps 1–3 · How it works

Three steps, one chat

Here the AI compares three proposals that each count success their own way. The prompt makes it use one measure and one cost basis first.

1
Open a new chat

Use the AI tool your company approves: ChatGPT, Claude, Gemini or Copilot.

2
Paste the prompt and the inputs

Copy the prompt and the five fictional inputs from the appendix into one message.

3
Recalculate the pilot yourself

$10.2M kept of $12.0M is 85%. Compare it with sales' $253.4M of $288M. Then decide what you take to the CEO.

The prompt, part by part
Copy-ready version in the appendix

I'm Chief Revenue Officer at an industrial automation company that sells annual service contracts and software subscriptions. Three leaders have each proposed a different way to own renewals: sales keeps them, customer success takes them, or a dedicated renewals team. I have to recommend one to our CEO and CFO on Thursday, March 11, with what it costs. Today is Tuesday, March 2. Below are the CEO's email, the three proposals, last year's renewal data, the pay notes and a note from our VP Customer Success.

1
Who you are, who you recommend to, the meeting date and today's date.

Please give me:

  1. A like-for-like table of the three options: what each claim measures (contracts or dollars), the period and accounts it covers, and the full yearly cost, including pay changes, for 2027 and from 2028.
  2. Each option tested against the data.
  3. What's missing that I'd need to decide well.
  4. My recommendation, including whether any option is proven well enough to choose now, and what must be confirmed, by whom, before March 11.
  5. A five-line summary for the CEO and CFO.
2
A like-for-like table with full costs for 2027 and from 2028, each option tested, what's missing, a recommendation with who confirms what, and a five-line summary.

Put every option on the same measure, period and cost basis before you compare them, and say plainly if none is proven as proposed. Check the arithmetic. Don't add facts that aren't in the inputs; if you assume something, say so.

3
The heavy lifting: put every option on the same measure, period and cost basis, and say plainly if none is proven. It found seven of eight planted problems and part of the eighth.
Where to paste
Every tool has one message box. Paste the prompt, add a blank line, paste the inputs, send. Renewal exports can be attached as files if your tool allows it.
Safe to try
Practise with the fictional inputs first. Real renewal data, pay plans and customer names are confidential: use only your company's approved tool.
The AI's first answer

What Claude actually wrote

Exactly as it came back, not edited. Read it the way Helena did: check each point against the source it cites.

UNEDITED
Claude · fresh session, no other context · Nov 15, 2026 · one run

Renewal ownership: three options compared on one basis

Basis used throughout (my assumptions, please confirm):

  • "Retention" means dollar retention: renewed value divided by the value up for renewal. Lost contracts and downgrades both count against it. The CEO's 88% matches this measure.
  • Costs are full-year renewal pay, calculated on 2026 renewed value of $263.6M. That is $253.4M from sales plus $10.2M from the pilot. I assume today's 2% account executive (AE) commission is paid on all of that value. If the pilot contracts were excluded, today's cost would be $5.07M, not $5.27M.
  • Salaries are as stated. Benefits, recruiting, ramp-up time and management cost are not in the inputs, so they are not included.
  • "2027" means a full year. The decision is made in March, so 2027 has already started. If the change starts mid-year, the 2027 figures would be prorated.

1. Like-for-like table

Retention, both ways (from the 2026 data):

Group Contracts renewed Dollars renewed
Sales-owned (1,100 contracts, $288M) 1,001 of 1,100 = 91.0% $253.4M of $288M = 88.0%
CS pilot (100 contracts, $12.0M) 95 of 100 = 95.0% $10.2M of $12.0M = 85.0%
Whole company (1,200 contracts, $300M) 1,096 of 1,200 = 91.3% $263.6M of $300M = 87.9%

The pilot's arithmetic checks out: $12.0M − $0.6M lost − $1.2M downgraded = $10.2M.

Options. Today's renewal pay is 2% × $263.6M = $5.27M a year.

A. Sales keeps renewals B. Customer success takes them C. Dedicated renewals team
Claim 88% → 92% "95% vs sales' 88%" 94%
What the claim measures Probably dollars (88% matches sales' dollar rate). Measure for 92% not stated. Mixed. 95% is the pilot's contract rate; 88% is sales' dollar rate. Not stated (vendor report) found
Period and accounts A forecast. No period or basis given. 100 contracts, each under $250K, due July to December 2026 only. None of the 60 largest. Industry benchmark, not our accounts. Period and sample unknown.
Like-for-like figure No evidence 85% in dollars vs sales' 88% (or 95% vs 91% in contracts) None from our data
Cost as stated ~$1.3M $1.12M $1.5M
Full yearly cost, 2027 2.5% × 263.6 = $6.59M (+$1.32M vs today) AE 2% $5.27M + CSM 1% $2.64M + 8 × $140K $1.12M = $9.03M (+$3.76M) AE 2% $5.27M + renewal managers 1.5% $3.95M + 10 × $150K $1.50M = $10.73M (+$5.45M) found
Full yearly cost, from 2028 $6.59M (+$1.32M) $2.64M + $1.12M = $3.76M (−$1.52M vs today) $3.95M + $1.50M = $5.45M (+$0.18M)

Notes on cost:

  • Option A's "about $1.3M" is the extra cost at today's renewed value. If retention really reached 92% of $300M ($276M renewed), commission would be $6.90M, which is +$1.63M.
  • Options B and C state salaries only. Neither includes its own variable pay. Neither includes the AEs' 2% that continues through 2027. Their real 2027 extra cost is about 3.4 times what was stated for B and 3.6 times for C. found
  • For scale: 1 point of dollar retention on a $300M base is $3M of contract value. Margin isn't in the inputs, so I can't turn that into profit.

2. Each option tested against the data

A. Sales keeps renewals, 2.5% commission.

  • The cost is consistent with the data.
  • The 92% has no support in the inputs. There's no history linking commission rate to retention, and no explanation of why 0.5 points of commission would add 4 points of retention. found
  • Not proven. It's the only option with no transition risk, and the easiest to reverse.

B. Customer success takes renewals.

  • The headline compares a contract rate with a dollar rate. On the same measure, the pilot did worse in dollars (85% vs 88%) and better in contracts (95% vs 91%).
  • Downgrades ($1.2M) were two-thirds of the pilot's dollar loss.
  • The pilot also differs from the sales book in three ways: found
    • Smaller contracts: about $120K average, vs about $262K in the sales book.
    • Half a year only: renewals due July to December.
    • No large accounts: the 60 largest are $132M, 44% of the base.
  • So it is not a like-for-like comparison even on dollars.
  • Priya's note on customer preference is anecdotal. No survey or data is given. The pilot's numbers do not "prove it." partly
  • Capacity: 8 CSMs would cover 1,200 contracts, about 150 each. The inputs don't say how many people ran the pilot's 100.
  • Not proven, and the dollar result points the other way.

C. Dedicated renewals team.

  • The only evidence is a vendor's industry benchmark. The measure, sample, period and the vendor's interest are all unknown. found
  • It's the costliest option in 2027 ($10.73M) and still costs slightly more than today from 2028.
  • Capacity: 10 managers for 1,200 contracts, about 120 each.
  • Not proven.

3. What's missing

  1. Sales' renewal rate for the pilot's segment: contracts under $250K due July to December 2026, in both contracts and dollars. This is the only fair comparison with the pilot.
  2. Sales' losses split into lost contracts and downgrades. For the pilot, we have that split. For sales' $34.6M shortfall, we don't.
  3. The basis for Mark's 92%, and whether it means dollars or contracts.
  4. The vendor report: source, how retention is defined, sample, date, and whether the vendor sells a renewals solution.
  5. Fully loaded cost per hire: benefits, recruiting, ramp-up time, managers. Also how long hiring takes, and who owns renewals in the meantime.
  6. Who owns the 60 largest contracts ($132M) under B and C, and the evidence that they would do as well as today. A: ask the customers
  7. Pilot staffing: how many CSMs worked the pilot's 100 contracts, and what drove the $1.2M of downgrades.
  8. The 2027 renewal base and forecast. I used 2026 renewed value as a stand-in.
  9. Whether AEs earn 2% on pilot contracts today, and on downgraded value.
  10. Profit margin on renewals, so cost can be weighed against retained revenue.

4. Recommendation

None of the three is proven as proposed. found

  • A's uplift is an unsupported forecast.
  • B's evidence, put on the same measure, shows lower dollar retention than sales.
  • C rests only on an outside vendor's benchmark.

What I'd recommend on March 11:

  • Don't approve B or C on the current evidence. Both are understated in cost: +$3.76M and +$5.45M in 2027, against the $1.12M and $1.5M stated.
  • Treat A as a known cost (+$1.3M to $1.6M a year) with an unknown benefit. Don't present 92% as an expected result.
  • Keep the decision open until the matched-segment comparison (item 1 above) is in. Then decide either which option to adopt or how to run a fair test, for example on matched accounts and the same measure. A test plan is my suggestion; it isn't in the inputs.
  • If the CEO needs a decision on the day, A is the least disruptive and most reversible, but it should be described as unproven. B: no change yet

To confirm before March 11 (ideally by Friday, March 5): C: three asks

  • Revenue operations: sales' renewal for the pilot's segment in contracts and dollars, sales' losses split into lost and downgraded, and the vendor report's details.
  • Mark Ellison: the basis and measure behind 92%.
  • Priya Nair: pilot staffing, the causes of the downgrades, and how B would cover the 60 largest accounts.
  • HR compensation and the CFO: fully loaded hire costs, hiring timeline, whether 2% is paid on pilot contracts today, and the 2027 renewal base. The CFO checks the cost table.

5. Five-line summary for the CEO and CFO

  1. Last year we kept 88% of renewal dollars (87.9% company-wide, 88.0% in sales-owned accounts) and 91% of contracts.
  2. None of the three proposals is proven. Sales' 92% has no stated basis; renewals-team 94% is a vendor benchmark; CS's "95% vs 88%" compares contracts with dollars.
  3. On the same dollar measure, the CS pilot kept 85% vs sales' 88%, on smaller, half-year contracts with none of our 60 largest ($132M).
  4. Full yearly renewal pay vs today's $5.27M:
    • A: $6.59M every year.
    • B: $9.03M in 2027, then $3.76M from 2028.
    • C: $10.73M in 2027, then $5.45M from 2028.
    • B and C as proposed understate their 2027 cost by $2.6M and $4.0M.
  5. Recommendation: don't approve B or C now. A is a known cost with an unknown benefit. Decide once revenue operations gives the matched-segment renewal rates and HR and the CFO confirm fully loaded costs.
foundcaught a planted problemA–CHelena changes it; see the next section
The human part

What Helena changed, and why

Claude put every claim on the same measure, found the pay each proposal left out and said none is proven. It doubted the customer claim but never asked for the customers' view. Helena's edits add it, take out the fallback, and cut the asks to three.

What Claude found: 7 of 8 planted problems, and part of the eighth
One run
✓Contracts compared with dollars
✓Pilot: six months, small contracts
✓B: pay left out, paid twice
✓C: pay left out
✓A's 92% unsupported
✓C's 94% is a vendor's
½Customers' view not asked for
✓None proven
Three edits, one catch and two judgment calls
A
CLAUDE WROTE
Ten missing items, and none asks what customers want. The customer claim is called “anecdotal” and left there.
HELENA'S VERSION
Helena adds what customers want, starting with the 60 largest contracts ($132M). Today it comes second-hand, from the VP who proposes option B.
WHY · WHAT'S MISSING
The decision is who the customer deals with at renewal. Doubting a second-hand claim isn't the same as finding out.
B
CLAUDE WROTE
“If the CEO needs a decision on the day, A is the least disruptive and most reversible, but it should be described as unproven.”
HELENA'S VERSION
If a decision is needed on March 11, Helena proposes no change for now: renewals stay with sales at today's 2% while revenue operations compares like with like.
WHY · DECISION
Option A is unproven too. Spending about $1.3M a year on an unsupported 92% is the same mistake at a smaller size.
C
CLAUDE WROTE
Ten missing items and four groups of owners to confirm before March 11.
HELENA'S VERSION
The page carries the asks that change the decision, by Friday, March 5: sales' results for the pilot's kind of contracts, fully loaded costs from HR and the CFO, and the customers' view.
WHY · LENGTH
The CEO and CFO need what decides the question. The rest stays on Helena's working list.
What only you can do
Ask the people the decision is about, keep your fallback as cautious as the evidence, and give the CEO only what decides it. The AI made the comparison fair. You decide what happens on March 11.
Your turn

Make it yours

Use it whenever several teams propose competing fixes for the same number: who owns renewals, which channel gets the budget, which team takes a segment.

Your prompt: fill in the brackets

I'm [your role] at a [industry] company. [Several teams] have each proposed a different way to [solve the problem]. I have to recommend one to [who] on [date], with what it costs. Today is [date]. Below are [their email, the proposals, the results data, the pay or cost notes and any other notes].

Please give me: 1. A like-for-like table of the options: what each claim measures, the period and accounts it covers, and the full yearly cost, including pay changes. 2. Each option tested against the data. 3. What's missing that I'd need to decide well. 4. My recommendation, including whether any option is proven well enough to choose now, and what must be confirmed, by whom, before [date]. 5. A five-line summary for [who].

Put every option on the same measure, period and cost basis before you compare them, and say plainly if none is proven as proposed. Check the arithmetic. Don't add facts that aren't in the inputs; if you assume something, say so.

Five checks before you recommend
1
Every percentage says what it counts, and they all count the same thing.
2
Every result covers comparable accounts and months.
3
Every cost includes the pay changes, in each year.
4
Outside benchmarks are labelled as outside.
5
The people affected, including customers, have been asked.
Make it repeatable (optional)
Save the prompt once and reuse it whenever proposals compete: a custom GPT in ChatGPT, a project in Claude, a Gem in Gemini, or an agent in Copilot. Some companies switch these off; copy and paste always works.
With real data
Renewal data, pay plans and customer names are confidential. Use only the AI tool your company approves, and follow its rules.
What this example shows
One run of this prompt in Claude found seven of eight problems planted in fictional inputs, and part of the eighth, and said plainly that none of the proposals is proven.
What it doesn't show
That every tool, setting or run will do the same, how much time it saves you, or who should own renewals at a real company. Talvora Group and its people are fictional.
Appendix

Copy-ready: the prompt and the five inputs

Copy the prompt, then all five inputs, into one message in a new chat. Everything here is fictional and safe to paste.

The prompt

I'm Chief Revenue Officer at an industrial automation company that sells annual service contracts and software subscriptions. Three leaders have each proposed a different way to own renewals: sales keeps them, customer success takes them, or a dedicated renewals team. I have to recommend one to our CEO and CFO on Thursday, March 11, with what it costs. Today is Tuesday, March 2. Below are the CEO's email, the three proposals, last year's renewal data, the pay notes and a note from our VP Customer Success.

Please give me:

  1. A like-for-like table of the three options: what each claim measures (contracts or dollars), the period and accounts it covers, and the full yearly cost, including pay changes, for 2027 and from 2028.
  2. Each option tested against the data.
  3. What's missing that I'd need to decide well.
  4. My recommendation, including whether any option is proven well enough to choose now, and what must be confirmed, by whom, before March 11.
  5. A five-line summary for the CEO and CFO.

Put every option on the same measure, period and cost basis before you compare them, and say plainly if none is proven as proposed. Check the arithmetic. Don't add facts that aren't in the inputs; if you assume something, say so.

[1] Email from Andrew Hale, CEO

Helena, renewals slipped to 88% last year and I'm getting three different answers on who should own them. Please bring Sofia and me one recommendation on March 11, with what it costs.

[2] The three proposals (one-line summaries from each author)
Option Proposed by Claim Cost stated
A. Sales keeps renewals, with a richer incentive Mark Ellison, VP Sales Renewal commission for account executives goes from 2% to 2.5% of renewed value. "Retention rises from 88% to 92%." About $1.3M a year
B. Customer success takes renewals Priya Nair, VP Customer Success "Our pilot retained 95% against sales' 88%." Hire 8 customer success managers (CSMs) at $140K each. $1.12M a year
C. A dedicated renewals team Revenue operations 10 renewal managers at $150K each. "Dedicated renewal teams achieve 94% (industry benchmark, vendor report)." $1.5M a year
[3] Renewal data, 2026 (revenue operations)
  • Renewal base: 1,200 annual contracts worth $300M. All renew once a year.
  • Sales-owned: 1,100 contracts, $288M. Renewed: 1,001 contracts and $253.4M.
  • Customer success pilot: 100 contracts, each under $250K a year, all due July to December 2026, $12.0M. Renewed: 95 contracts and $10.2M (5 lost, $0.6M; downgrades at renewal, $1.2M).
  • The 60 largest contracts (each over $1M) are $132M of the base. None was in the pilot.
[4] Pay notes (HR compensation)
  • Today account executives earn 2% of renewed value.
  • Option B: CSMs earn a 1% bonus on renewed value. Account executives keep their 2% for 2027 "so they support the hand-over"; it ends in 2028.
  • Option C: renewal managers earn 1.5% of renewed value. Account executives stop earning on renewals from January 2028.
[5] Note from Priya Nair, VP Customer Success

"Customers tell my team they'd rather renew with the people who help them every day than with a salesperson they see once a year. The pilot proves it."